Designing band-pass multisine excitations for microwave behavioral model identification

This work deals with the design of band-pass multisine excitations appropriate for microwave nonlinear model extraction and validation. Several examples show that the probability with which each instantaneous amplitude value is reached plays a determinant role on the response of a memoryless nonlinear system. Moreover, further tests made with nonlinear dynamic systems have shown that even the signal probability density function is insufficient, and one needs to know the signal's waveform, or the complete multisine amplitude and phase distribution.